Schema-Aware Multi-Task Learning for Complex Text-to-SQL

arXiv:2403.09706v1 Announce Type: new
Abstract: Conventional text-to-SQL parsers are not good at synthesizing complex SQL queries that involve multiple tables or columns, due to the challenges inherent in identifying the correct schema items and performing accurate alignment between question and schema items. To address the above issue, we present a schema-aware multi-task learning framework (named MTSQL) for complicated SQL queries. Specifically, we design a schema linking discriminator module to distinguish the valid question-schema linkings, which explicitly instructs the encoder by …

abstract alignment arxiv challenges cs.ai cs.cl cs.db framework good issue multiple multi-task learning queries question schema sql sql queries tables text text-to-sql type